The Rise of Low-Cost AI Disrupting Investment Structures - Is This the End of the Massive Infrastructure Bubble?
Cost: The Foundation of All Business
In business, the importance of cost is a universal principle. Whether launching a new venture or improving an existing one, return on investment is always considered. No matter how innovative an idea or technology may be, it cannot succeed as a business if the costs are not justified.
AI is no exception. In fact, the enormous cost structure of AI development has been a barrier to entry for many companies until now.
The True Nature of the AI Investment Bubble
AI development has required massive investments: high-performance chips, large-scale data centers, and enormous power consumption. This investment structure, considered “essential,” has been the basis for rising stock prices and investment frenzy in AI-related companies.
NVIDIA’s stock price rose approximately 240% in 2023 alone, and investments in data center construction reached record highs. Capital investments required for AI development resembled a gold rush.
The Shifting Paradigm of “Common Sense”
However, the emergence of DeepSeek-V3 suggests the potential to fundamentally overturn the premises of this investment structure. With a development cost of approximately 1 billion yen (an exceptional bargain), it achieved performance equal to or better than cutting-edge models. Furthermore, its dependence on expensive specialized chips has been kept low.
This fact suggests that the market’s conventional wisdom that “AI requires massive investment” is not necessarily absolute.
Impact on the Market
This change could significantly impact AI-related investments.
Chip manufacturers for AI development will likely be most directly affected. Reduced dependence on high-performance chips could impact the revenue forecasts of these companies.
Data center investments may also need reconsideration. If the adoption of efficient architectures reduces the required computational resources, the validity of large-scale infrastructure investments will be questioned.
For power companies, predictions of increased power demand due to AI development may need revision. The emergence of more efficient models could suppress the anticipated growth in power consumption.
The Future of the Investment Bubble
Such changes will likely affect the valuation of AI-related stocks. Stock price levels formed on the premise that “AI requires massive investment” may need reconsideration.
In particular, companies specializing in infrastructure provision may require more careful evaluation. Meanwhile, attention may increase for companies with new technologies and methods that enable efficient AI development.
New Investment Opportunities
However, these changes are not entirely negative. The possibility of low-cost AI development will also create new investment opportunities.
For example, the creation of new businesses utilizing AI technology will become more active. Additionally, companies that have hesitated to adopt AI due to cost concerns may accelerate their implementation.
Future Outlook
The AI industry is about to enter a new phase of “breaking away from high-cost structures.” While this change may cause temporary market disruption, it could lead to healthy market development in the long term.
For investors, it’s important to not be bound by previous “common sense” and to calmly assess new technological trends and changes in market structure. The AI revolution continues, but its form may be different from what we expect.